A control method and system for a direct-heated cathode electron gun based on image processing

Abstract

The present invention provides a control method and system for a direct-heated cathode electron gun based on image processing, and the present invention relates to the technical field of direct-heated cathode electron guns. The method comprises the following steps: obtaining an infrared camera video of a film material under an initial power of the electron gun; determining a plurality of electron gun test powers using a video processing model based on the infrared camera video of the film material under the initial power of the electron gun; obtaining an infrared camera video of the film material under a plurality of electron gun test powers; determining a plurality of recommended powers using a recommended power determination model based on the infrared camera video of the film material under the plurality of electron gun test powers; determining a target power based on the plurality of recommended powers; and controlling the electron gun based on the target power. The method can accurately determine a suitable electron gun power.

Description

Technical Field

[0001] The invention relates to the technical field of directly heated cathode electron guns, and in particular to a control method and system of a directly heated cathode electron gun based on image processing. Background Art

[0002] The direct-heated cathode electron gun is the core component of vacuum coating equipment and is widely used in the evaporation and coating processes of high-melting-point materials (such as tungsten and molybdenum) and active materials (such as titanium and niobium). The thermal electron beam is generated by heating the cathode filament. The electron beam bombards the target material and converts its thermal energy into evaporation heat energy, thereby achieving the evaporation and coating of the material. However, under high-power conditions, a large amount of gas will be released during the evaporation of the film material, causing part of the vaporized material in the cavity to splash in the form of micron-sized droplets. This splashing phenomenon may cause the molten film material to splash onto the electron gun filament, causing filament contamination, thereby shortening the filament life and even causing the filament to break. However, when the power of the electron gun is too low, the evaporation rate of the film material is insufficient, which will also lead to a longer coating time.

[0003] Therefore, how to accurately determine the appropriate power of the electron gun is a problem that needs to be solved urgently. Summary of the invention

[0004] The main technical problem solved by the present invention is how to accurately determine the power of a suitable electron gun.

[0005] According to a first aspect, the present invention provides a control method for a direct-heated cathode electron gun based on image processing, comprising: obtaining an infrared camera video of a film material at an initial power of the electron gun; determining a plurality of electron gun test powers using a video processing model based on the infrared camera video of the film material at the initial power of the electron gun; obtaining an infrared camera video of the film material at a plurality of electron gun test powers; determining a plurality of recommended powers using a recommended power determination model based on the infrared camera videos of the film material at the plurality of electron gun test powers; determining a target power based on the plurality of recommended powers; and controlling the electron gun based on the target power.

[0006] In one possible implementation, determining the target power based on the multiple recommended powers includes: using a generative adversarial network to generate an infrared simulation video of the film material at each electron gun recommended power based on the multiple recommended powers and the infrared camera videos of the film material at the multiple electron gun test powers; constructing a knowledge graph, the knowledge graph including multiple recommended power nodes and multiple edges between the multiple recommended power nodes, each recommended power node being a recommended power, a node feature of each recommended power node including an infrared simulation video of the film material at an electron gun recommended power, and an edge between two recommended power nodes representing a power difference between the two recommended powers; and determining the target power by processing the knowledge graph based on a graph convolutional network.

[0007] In a possible implementation, the video processing model is a Transformer model.

[0008] In a possible implementation manner, the recommended power determination model is a Transformer model.

[0009] According to a second aspect, the present invention provides a control system for a direct-heated cathode electron gun based on image processing, comprising:

[0010] A first acquisition module is used to acquire infrared camera video of the film material under the initial power of the electron gun;

[0011] A test power determination module, used to determine a plurality of electron gun test powers using a video processing model based on an infrared camera video of the film material at the initial power of the electron gun;

[0012] A second acquisition module is used to acquire infrared camera videos of the film material under multiple electron gun test powers;

[0013] A recommended power determination module, configured to determine a plurality of recommended powers using a recommended power determination model based on infrared camera videos of the film material under the plurality of electron gun test powers;

[0014] A target power determination module, configured to determine a target power based on the multiple recommended powers;

[0015] A control module is used to control the electron gun based on the target power.

[0016] In a possible implementation manner, the target power determination module is further configured to:

[0017] Generate an infrared simulation video of the film material at each electron gun recommended power using a generative adversarial network based on the infrared camera videos of the film material at the multiple recommended powers and the multiple electron gun test powers;

[0018] Constructing a knowledge graph, wherein the knowledge graph includes a plurality of recommended power nodes and a plurality of edges between the plurality of recommended power nodes, each recommended power node is a recommended power, a node feature of each recommended power node includes an infrared simulation video of a film material under an electron gun recommended power, and an edge between two recommended power nodes represents a power difference between the two recommended powers;

[0019] The knowledge graph is processed based on the graph convolutional network to determine the target power.

[0020] In a possible implementation, the video processing model is a Transformer model.

[0021] In a possible implementation manner, the recommended power determination model is a Transformer model.

[0022] According to a third aspect, an embodiment of the present invention provides an electronic device, comprising: a processor; a memory; and a computer program; wherein the computer program is stored in the memory and is configured to be executed by the processor to implement a method as described above, the method comprising: obtaining an infrared camera video of a film material under an initial power of an electron gun; determining a plurality of electron gun test powers using a video processing model based on the infrared camera video of the film material under the initial power of the electron gun; obtaining an infrared camera video of the film material under a plurality of electron gun test powers; determining a plurality of recommended powers using a recommended power determination model based on the infrared camera videos of the film material under the plurality of electron gun test powers; determining a target power based on the plurality of recommended powers; and controlling the electron gun based on the target power.

[0023] According to the fourth aspect, the present embodiment provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned control method for a direct-heated cathode electron gun based on image processing, the method comprising: obtaining an infrared camera video of a film material at an initial power of the electron gun; determining a plurality of electron gun test powers using a video processing model based on the infrared camera video of the film material at the initial power of the electron gun; obtaining an infrared camera video of the film material at a plurality of electron gun test powers; determining a plurality of recommended powers using a recommended power determination model based on the infrared camera videos of the film material at the plurality of electron gun test powers; determining a target power based on the plurality of recommended powers; and controlling the electron gun based on the target power.

[0024] The present invention provides a control method and system for a direct-heated cathode electron gun based on image processing. The method comprises obtaining an infrared camera video of a film material under an initial power of the electron gun; determining a plurality of electron gun test powers using a video processing model based on the infrared camera video of the film material under the initial power of the electron gun; obtaining an infrared camera video of the film material under a plurality of electron gun test powers; determining a plurality of recommended powers using a recommended power determination model based on the infrared camera video of the film material under the plurality of electron gun test powers; determining a target power based on the plurality of recommended powers; and controlling the electron gun based on the target power. The method can accurately determine a suitable electron gun power. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A schematic diagram of an application scenario of a control method for a direct-heated cathode electron gun based on image processing provided by an embodiment of the present invention;

[0026] Figure 2A schematic flow chart of a control method of a direct-heated cathode electron gun based on image processing provided by an embodiment of the present invention;

[0027] Figure 3 A schematic diagram of a process for determining a target power based on multiple recommended powers provided by an embodiment of the present invention;

[0028] Figure 4 A schematic diagram of a control system of a direct-heated cathode electron gun based on image processing provided by an embodiment of the present invention;

[0029] Figure 5 A schematic diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0030] The present invention is further described in detail below by specific embodiments in conjunction with the accompanying drawings. Wherein similar elements in different embodiments adopt associated similar element numbers. In the following embodiments, many detailed descriptions are for making the present invention better understood. However, those skilled in the art can easily recognize that some features can be omitted in different situations, or can be replaced by other elements, materials, methods. In some cases, some operations related to the present invention are not shown or described in the specification, this is to avoid the core part of the present invention being overwhelmed by too much description, and for those skilled in the art, it is not necessary to describe these related operations in detail, and they can fully understand the related operations according to the description in the specification and the general technical knowledge in the art.

[0031] Figure 1 A schematic diagram of an application scenario of a control method for a directly heated cathode electron gun based on image processing provided in an embodiment of the present invention. Figure 1 The application scenario of the control method of the direct-heated cathode electron gun based on image processing may include a server 11, a network 12, a terminal 13 and a storage device 14.

[0032] In some embodiments, the server 11 may be a single server or a server group. The server 11 may access information and / or data stored in the terminal 13 or the storage device 14 through the network 12. In some embodiments, the server 11 may be used to execute Figure 2 The control method of the direct-heated cathode electron gun based on image processing is shown in.

[0033] The network 12 may facilitate the exchange of information and / or data. In some embodiments, the network 12 may be any form of wired or wireless network, or any combination thereof.

[0034] Terminal 13 may refer to one or more terminal devices used by a user. In some embodiments, terminal 13 may include one or more combinations of a mobile device, a tablet computer, a laptop computer, and the like.

[0035] The storage device 14 may store data and / or instructions. For example, the storage device 14 may store data instructions of a control method of a direct-heated cathode electron gun based on image processing.

[0036] In an embodiment of the present invention, there is provided Figure 2 A control method for a direct-heated cathode electron gun based on image processing is shown, and the control method for a direct-heated cathode electron gun based on image processing includes steps S1 to S6:

[0037] Step S1, obtaining an infrared camera video of the film material under the initial power of the electron gun.

[0038] The electron gun initial power is the default power setting of the electron gun at the beginning of the coating process. It is usually a lower basic power value used for preliminary heating of the film material.

[0039] The film material is the material to be evaporated, usually a high melting point metal (such as tungsten, molybdenum) or a highly active material (such as titanium, niobium).

[0040] Infrared camera video is a video shot by infrared thermal imaging equipment to capture the dynamic changes of the film material during the heating process. Infrared camera video can show whether the film material splashes onto the electron gun filament.

[0041] Step S2, determining a plurality of electron gun test powers using a video processing model based on the infrared camera video of the film material at the initial power of the electron gun.

[0042] The video processing model is a Transformer model. The Transformer model includes an encoder and a decoder. The encoder is used to represent the input sequence, which includes a self-attention mechanism and a feed-forward network. Based on the encoder, the decoder introduces an additional multi-head attention mechanism to decode the encoder output and generate a target sequence. The Transformer model can be used to process infrared camera videos of film materials at the initial power of the electron gun, and can better capture the relationship in the time series of the video.

[0043] The initial power is not enough to meet the coating requirements, but too high a power may cause the film material to splash. By analyzing the infrared video through the video processing model, multiple electron gun test powers can be determined to avoid the inefficiency caused by blindly setting the test power.

[0044] The infrared camera video records the dynamic information of the film material such as temperature distribution, thermal radiation intensity and possible droplet splash trajectory under the initial power of the electron gun. In infrared video analysis, Transformer can simultaneously focus on the temporal dependency between video frames (such as temperature change trend) and the spatial dependency within the frame, and finally determine multiple electron gun test powers.

[0045] Step S3, obtaining infrared camera videos of the film material under multiple electron gun test powers.

[0046] After a plurality of electron gun test powers are determined, the power of the electron gun is adjusted to each test power, and infrared camera videos of the film material under the plurality of electron gun test powers are obtained.

[0047] The infrared video of the film material under multiple electron gun test powers is the infrared video of the film material shot by an infrared camera under different test powers.

[0048] Step S4, determining a plurality of recommended powers using a recommended power determination model based on the infrared camera videos of the film material under the plurality of electron gun test powers.

[0049] The recommended power determination model is a Transformer model. The infrared videos at different test powers provide evaporation characteristic data of the film material at different powers, and the power setting can be further optimized through the recommended power determination model.

[0050] The infrared videos at different test powers provide the evaporation characteristic data of the film material under different conditions. These data provide rich experimental basis for subsequent power optimization.

[0051] The test power may still be too wide and further optimization is needed to determine the final power value. The recommended power is a better power selected based on the test data, which can improve the scientificity and accuracy of the power setting.

[0052] Infrared video can reflect the evaporation rate of the film material at different powers. Too fast evaporation rate may cause splashing, while too slow evaporation rate may lead to low coating efficiency.

[0053] Infrared video can capture the temperature distribution on the film surface and determine whether there is local overheating or uneven heating.

[0054] Under high power, the film material may splash due to bubble bursting or pressure difference, and infrared video can capture these dynamic changes.

[0055] In some embodiments, the recommended power determination model includes a feature extraction layer, an optimal test power output layer, and a recommended power determination layer. The feature extraction layer, the optimal test power output layer, and the recommended power determination layer all include a Transformer structure. The input of the feature extraction layer is an infrared camera video of the film material under multiple electron gun test powers, the output of the feature extraction layer is the test information of the film material under each electron gun test power, the input of the optimal test power output layer is the test information of the film material under each electron gun test power, the output of the optimal test power output layer is the optimal test power, the input of the recommended power determination layer is the optimal test power and the test information of the film material under each electron gun test power, and the output of the recommended power determination layer is multiple recommended powers.

[0056] The feature extraction layer focuses on extracting key film material evaporation characteristics from infrared videos, providing high-quality feature data for subsequent layers. The optimal test power output layer predicts the optimal evaporation characteristics at each test power based on the extracted features. The recommended power determination layer combines the optimal test power and feature information to ultimately determine multiple recommended powers.

[0057] The test information of the film material under each electron gun test power includes droplet splash trajectory sequence information, film material evaporation bubble burst sequence information, and molten film material surface change sequence information.

[0058] The droplet splash trajectory sequence information refers to the dynamic trajectory information of droplet splash formed by the molten film material due to bubble bursting or pressure difference during the evaporation process of the film material.

[0059] The film material evaporation bubble rupture sequence information refers to the dynamic information of the bubbles formed by the release of internal gas rupturing on the film material surface during the film material heating process.

[0060] The surface change sequence information of the molten film material refers to the dynamic change information of the film material surface due to temperature change and gas release during the heating process of the film material.

[0061] Step S5: determining a target power based on the multiple recommended powers.

[0062] In some embodiments, Figure 3 A schematic diagram of a process for determining a target power based on multiple recommended powers is provided in an embodiment of the present invention, wherein determining the target power based on multiple recommended powers includes steps S21 to S23:

[0063] Step S21, based on the infrared camera videos of the film material at the multiple recommended powers and the multiple electron gun test powers, using a generative adversarial network to generate an infrared simulation video of the film material at each electron gun recommended power.

[0064] The infrared simulation video of the film material at each recommended electron gun power is an infrared simulation video generated by a generative adversarial network.

[0065] The Generative Adversarial Network (GAN) includes a generator and a discriminator, which compete with each other and continuously optimize to achieve the purpose of generating realistic data. The GAN can learn the feature representation of the data through the training process. The input of the GAN is the infrared camera video of the film material under the multiple recommended powers and the multiple electron gun test powers, and the output of the GAN is the infrared simulation video of the film material under each electron gun recommended power.

[0066] By generating simulated videos, we can reduce the reliance on actual infrared camera videos, save time and computing resources, and avoid manually adjusting the power and obtaining videos for each power. Since a large amount of video data is required to determine the final target power in the actual process, it may be difficult to capture the evaporation characteristics of the film material at all recommended powers during the actual shooting process, especially in some extreme cases. Generative adversarial networks can simulate these complex scenarios and provide more comprehensive training data. Generative adversarial networks are used to generate infrared simulation videos of the film material at each recommended power of the electron gun, which improves the efficiency of determining the target power. Generative adversarial networks are used to generate infrared simulation videos of the film material at each recommended power, so that more efficient and convenient analysis can be performed in the subsequent graph neural network model.

[0067] The generative adversarial network can learn the feature representation in the infrared camera video through the training process. The generator can generate realistic simulated videos based on the input recommended power and real video, while the discriminator is responsible for distinguishing the real video from the generated simulated video. The generative adversarial network can capture the complex relationship between the evaporation characteristics of the film material under different recommended powers. Thus, a realistic infrared simulated video of the film material under each recommended power of the electron gun is generated.

[0068] Step S22, constructing a knowledge graph, wherein the knowledge graph includes multiple recommended power nodes and multiple edges between the multiple recommended power nodes, each recommended power node is a recommended power, and the node feature of each recommended power node includes an infrared simulation video of the film material under the recommended power of an electron gun, and the edge between two recommended power nodes represents the power difference between the two recommended powers.

[0069] The knowledge graph is a database that represents knowledge in a graph structure, including nodes and edges. The node representing each recommended power in the knowledge graph is a recommended power node. The node feature of each recommended power node includes an infrared simulation video of the film material under the recommended power of the electron gun.

[0070] The edge between two recommended power nodes represents the power difference between the two recommended powers.

[0071] Step S23, processing the knowledge graph based on the graph convolutional network to determine the target power.

[0072] Graph Convolutional Network (GCN) is a deep learning model for processing graph data. It can process knowledge graphs, analyze the relationship between recommended power nodes, and ultimately determine the optimal target power.

[0073] The knowledge graph represents knowledge in a graphical structure, which can intuitively display the recommended powers and their relationships. Each recommended power node contains an infrared simulation video of the film material under the recommended power of the electron gun, and the edges between the nodes represent the power difference. This structured representation makes it easier for the graph convolutional network to understand and process complex relationships. Through the information of node features and edges, the graph convolutional network can better capture the similarities and differences between the recommended powers. For example, the infrared simulation video in the node features provides the evaporation characteristics of the film material at different powers, while the power difference on the edge reflects the changing relationship between these characteristics. The construction of the knowledge graph enables the recommended powers and their relationships to be effectively organized, which is convenient for subsequent graph convolutional networks to process. This organization method can reduce computational complexity and improve the training and reasoning efficiency of the model.

[0074] Graph convolutional networks (GCNs) can extract features and propagate information from nodes in knowledge graphs, effectively capturing the complex relationship between recommendation powers. Through convolution operations, graph convolutional networks can learn the feature representation of nodes and perform relational reasoning based on edge information to determine the optimal target power.

[0075] Since the knowledge graph provides structured node and edge information, the graph convolutional network can better process this information and avoid the problems of data sparsity and dimensionality disaster in traditional methods. This structural advantage makes GCN more advantageous in dealing with complex relationships.

[0076] By comprehensively analyzing the characteristics and relationships of the recommended power nodes, the graph convolutional network can find the optimal power that can both meet the coating requirements and avoid splashing of the film material onto the electron gun filament.

[0077] Based on the same inventive concept, Figure 4 A schematic diagram of a control system of a direct-heated cathode electron gun based on image processing provided by an embodiment of the present invention, wherein the control system of the direct-heated cathode electron gun based on image processing comprises:

[0078] The first acquisition module 41 is used to acquire the infrared camera video of the film material under the initial power of the electron gun;

[0079] A test power determination module 42, configured to determine a plurality of electron gun test powers using a video processing model based on an infrared camera video of the film material at the electron gun initial power;

[0080] The second acquisition module 43 is used to acquire infrared camera videos of the film material under multiple electron gun test powers;

[0081] A recommended power determination module 44 is used to determine a plurality of recommended powers using a recommended power determination model based on the infrared camera videos of the film material under the plurality of electron gun test powers;

[0082] A target power determination module 45, configured to determine a target power based on the plurality of recommended powers;

[0083] The control module 46 is configured to control the electron gun based on the target power.

[0084] Based on the same inventive concept, an embodiment of the present invention provides an electronic device, such as Figure 5 As shown, including:

[0085] The invention comprises: a processor 51; a memory 52; and a computer program; wherein the computer program is stored in the memory 52 and is configured to be executed by the processor 51 to implement the control method of a direct-heated cathode electron gun based on image processing as provided above, the method comprising: obtaining an infrared camera video of a film material at an initial power of the electron gun; determining a plurality of electron gun test powers using a video processing model based on the infrared camera video of the film material at the initial power of the electron gun; obtaining an infrared camera video of the film material at a plurality of electron gun test powers; determining a plurality of recommended powers using a recommended power determination model based on the infrared camera videos of the film material at the plurality of electron gun test powers; determining a target power based on the plurality of recommended powers; and controlling the electron gun based on the target power.

[0086] Based on the same inventive concept, this embodiment provides a computer-readable storage medium having a computer program stored thereon, which, when executed by the processor 51, implements the aforementioned control method for a direct-heated cathode electron gun based on image processing, the method comprising: obtaining an infrared camera video of a film material at an initial power of the electron gun; determining a plurality of electron gun test powers using a video processing model based on the infrared camera video of the film material at the initial power of the electron gun; obtaining an infrared camera video of the film material at a plurality of electron gun test powers; determining a plurality of recommended powers using a recommended power determination model based on the infrared camera videos of the film material at the plurality of electron gun test powers; determining a target power based on the plurality of recommended powers; and controlling the electron gun based on the target power.

[0087] The control method of the direct-heated cathode electron gun based on image processing provided in the embodiment of the present application can be applied to terminal devices (such as mobile phones), tablet computers, laptops, ultra-mobile personal computers (ultra-mobile personal computers, UMPCs), handheld computers, netbooks, personal digital assistants (personal digital assistants, PDAs), wearable devices (such as smart watches, smart glasses or smart helmets, etc.), augmented reality (augmented reality, AR) \ virtual reality (virtual reality, VR) devices, smart home devices, car computers and other electronic devices, and the embodiment of the present application does not impose any restrictions on this.

[0088] Similarly, it should be noted that in order to simplify the description disclosed in this specification and thus help understand one or more embodiments of the invention, in the above description of the embodiments of this specification, multiple features are sometimes combined into one embodiment, figure or description thereof. However, this disclosure method does not mean that the features required by the subject matter of this specification are more than the features mentioned in the claims. In fact, the features of the embodiments are less than all the features of the single embodiment disclosed above.

[0089] Finally, it should be understood that the embodiments described in this specification are only used to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, as an example and not a limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly introduced and described in this specification.

Claims

1. A control method for a direct-heated cathode electron gun based on image processing, characterized in that: include: Obtain infrared video of the film material under the initial power of the electron gun; Determine multiple electron gun test powers using a video processing model based on the infrared camera video of the film material at the initial power of the electron gun; Obtain infrared camera videos of film materials under multiple electron gun test powers; Based on the infrared camera videos of the film material under the multiple electron gun test powers, a recommended power determination model is used to determine multiple recommended powers, the recommended power determination model includes a feature extraction layer, an optimal test power output layer, and a recommended power determination layer, the feature extraction layer, the optimal test power output layer, and the recommended power determination layer all include a Transformer structure, the input of the feature extraction layer is the infrared camera videos of the film material under the multiple electron gun test powers, the output of the feature extraction layer is the test information of the film material under each electron gun test power, the input of the optimal test power output layer is the test information of the film material under each electron gun test power, the output of the optimal test power output layer is the optimal test power, the input of the recommended power determination layer is the optimal test power and the test information of the film material under each electron gun test power, and the output of the recommended power determination layer is multiple recommended powers; determining a target power based on the plurality of recommended powers; The electron gun is controlled based on the target power.

2. The control method of the direct-heated cathode electron gun based on image processing according to claim 1, characterized in that: The determining the target power based on the multiple recommended powers comprises: Generate an infrared simulation video of the film material at each electron gun recommended power using a generative adversarial network based on the infrared camera videos of the film material at the multiple recommended powers and the multiple electron gun test powers; Constructing a knowledge graph, wherein the knowledge graph includes a plurality of recommended power nodes and a plurality of edges between the plurality of recommended power nodes, each recommended power node is a recommended power, a node feature of each recommended power node includes an infrared simulation video of a film material under an electron gun recommended power, and an edge between two recommended power nodes represents a power difference between the two recommended powers; The knowledge graph is processed based on the graph convolutional network to determine the target power.

3. The control method of the direct-heated cathode electron gun based on image processing according to claim 1, characterized in that: The video processing model is a Transformer model.

4. The control method of the direct-heated cathode electron gun based on image processing according to claim 1, characterized in that: The recommended power determination model is a Transformer model.

5. A control system for a direct-heated cathode electron gun based on image processing, characterized in that: include: A first acquisition module is used to acquire infrared camera video of the film material under the initial power of the electron gun; A test power determination module, used to determine a plurality of electron gun test powers using a video processing model based on an infrared camera video of the film material at the initial power of the electron gun; A second acquisition module is used to acquire infrared camera videos of the film material under multiple electron gun test powers; A recommended power determination module, for determining a plurality of recommended powers using a recommended power determination model based on the infrared camera videos of the film material under the plurality of electron gun test powers, wherein the recommended power determination model comprises a feature extraction layer, an optimal test power output layer, and a recommended power determination layer, wherein the feature extraction layer, the optimal test power output layer, and the recommended power determination layer all comprise a Transformer structure, wherein the input of the feature extraction layer is the infrared camera videos of the film material under the plurality of electron gun test powers, the output of the feature extraction layer is the test information of the film material under each electron gun test power, the input of the optimal test power output layer is the test information of the film material under each electron gun test power, the output of the optimal test power output layer is the optimal test power, the input of the recommended power determination layer is the optimal test power and the test information of the film material under each electron gun test power, and the output of the recommended power determination layer is a plurality of recommended powers; A target power determination module, configured to determine a target power based on the multiple recommended powers; A control module is used to control the electron gun based on the target power.

6. The control system of the direct-heated cathode electron gun based on image processing according to claim 5, characterized in that: The target power determination module is also used for: Generate an infrared simulation video of the film material at each electron gun recommended power using a generative adversarial network based on the infrared camera videos of the film material at the multiple recommended powers and the multiple electron gun test powers; Constructing a knowledge graph, wherein the knowledge graph includes a plurality of recommended power nodes and a plurality of edges between the plurality of recommended power nodes, each recommended power node is a recommended power, a node feature of each recommended power node includes an infrared simulation video of a film material under an electron gun recommended power, and an edge between two recommended power nodes represents a power difference between the two recommended powers; The knowledge graph is processed based on the graph convolutional network to determine the target power.

7. The control system of the direct-heated cathode electron gun based on image processing according to claim 5, characterized in that: The video processing model is a Transformer model.

8. The control system of the direct-heated cathode electron gun based on image processing according to claim 5, characterized in that: The recommended power determination model is a Transformer model.

9. An electronic device, characterized in that: include: processor; Memory; and a computer program; wherein the computer program is stored in the memory and is configured to be executed by the processor to implement the control method of the direct-heated cathode electron gun based on image processing as claimed in any one of claims 1 to 4.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the control method of a direct-heated cathode electron gun based on image processing as claimed in any one of claims 1 to 4 is implemented.

Images